API tools faq
paste
Advertisement
Guest User

Untitled

a guest
May 28th, 2016
353
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Lua 10.67 KB | None | 0 0
raw download clone embed print report
  1. -- Parsed as per W3's HTML Standard
  2.  
  3. return function(HTMLSource)
  4.    
  5.    
  6.     local PARSER_POSITION = 1
  7.     local DATA_STATE = "" -- State of byte consumption
  8.     local TEXT_COLLECTION -- Used to collect character tokens. Instead of emitting them as a single character
  9.     local DOCUMENT = {} -- The final, tokenized, document structure.
  10.     local OPEN_ELEMENTS = {} -- Track elements that have not been closed. Used to determine parents and children. (If a tag is open, and then another is opened, we've found a family)
  11.     local TEXT_COLLECTION = "" -- Collection of unprocessed unicode. Dumped into a textnode if left unused without an opening tag encountered.
  12.     local ACTIVE_TOKEN -- Reference to the currently active token
  13.     local ACTIVE_ATTRIBUTE -- Reference to the currently active attribute
  14.     local FLAGS = { -- UNUSED
  15.         PAUSE_PARSER = false;
  16.         INTAG = false
  17.     }
  18.     local TOKENS = {}
  19.     local VALID_TAGS = {
  20.         ['scrollingframe'] = "ScrollingFrame";
  21.         ['frame'] = "Frame";
  22.         ['imagebutton'] = "ImageButton";
  23.         ['imagelabel'] = "ImageLabel";
  24.         ['textlabel'] = "TextLabel";
  25.         ['textbox'] = "TextBox";
  26.         ['textbutton'] = "TextButton";
  27.     }
  28.    
  29.     local function CreateToken(Type)
  30.         local Token = {
  31.             TagName = Type;
  32.             SelfClosingFlag = false;
  33.             IsClosingTag = false; -- Key element to making the tree.
  34.             Data = "";
  35.             Attributes = {};
  36.             Style = {};
  37.         }
  38.         table.insert(OPEN_ELEMENTS, Token)
  39.         table.insert(TOKENS, Token)
  40.         if (ACTIVE_TOKEN) then
  41.             -- I believe this is an impossible parser error. Never had it happen.
  42.             if (FindValInTable(OPEN_ELEMENTS, ACTIVE_TOKEN)) then
  43.                 table.insert(ACTIVE_TOKEN.Children, Token)
  44.             end
  45.         end
  46.         return Token
  47.     end
  48.    
  49.     local function AddAttribute(TOKEN)
  50.         local Attr = {AttributeName = "", Value=""}
  51.         table.insert(TOKEN.Attributes, Attr)
  52.         return Attr
  53.     end
  54.    
  55.     function FindValInTable(Tab, Val)
  56.         for i,v in pairs(Tab) do
  57.             if (v == Val) then
  58.                 return v, i
  59.             end
  60.         end
  61.     end
  62.    
  63.     local function GetLastOpenTag()
  64.         -- UNUSED
  65.         -- Searches for the most recently opened tag in OPEN_ELEMENTS
  66.         if (#OPEN_ELEMENTS > 0) then
  67.             return OPEN_ELEMENTS[#OPEN_ELEMENTS]
  68.         end
  69.     end
  70.    
  71.     local function EmitCharToOpenToken(Character)
  72.         local OpenToken = ACTIVE_TOKEN
  73.         OpenToken.Data = OpenToken.Data .. Character
  74.     end
  75.    
  76.     local function EmitToken(TOKEN)
  77.         -- Why do we add the token to the DOCUMENT without knowing its children?
  78.         -- Because closing tags are ALSO added to the document.
  79.         -- The closing tags will tell us when a nest ends.
  80.         -- EXAMPLE:
  81.         --[[
  82.             OPEN TAG A
  83.             OPEN TAG B
  84.             CLOSE TAG B
  85.             CLOSE TAG A
  86.         ]]
  87.         -- Because closing tags are there, we will know when one nest ends. We know OPEN TAG B is inside of A because it comes after A and has a valid close tag.
  88.        
  89.         -- Verify the tag name is a valid tag from VALID_TAGS
  90.         -- You may remove this check to accept arbitrary HTML tag names not supported via ROBLOX GUIs
  91.         if (VALID_TAGS[TOKEN.TagName] or TOKEN.TagName:lower() == "comment" or TOKEN.TagName:lower() == "__text") then
  92.             if (VALID_TAGS[TOKEN.TagName]) then
  93.                 TOKEN.TagName = VALID_TAGS[TOKEN.TagName]
  94.             end
  95.             table.insert(DOCUMENT, ACTIVE_TOKEN)
  96.             TOKENS = RemoveValueFromTable(TOKENS, ACTIVE_TOKEN)
  97.             if (TOKEN.SelfClosingFlag) then
  98.                 -- If it is self-closing, remove it from the open list. Otherwise, it is still open.
  99.                 OPEN_ELEMENTS = RemoveValueFromTable(OPEN_ELEMENTS, ACTIVE_TOKEN)
  100.             end
  101.         end
  102.         ACTIVE_TOKEN = nil--GetLastOpenTag()
  103.         ACTIVE_ATTRIBUTE = nil
  104.     end
  105.        
  106.    
  107.     function RemoveValueFromTable(Tab, Val)
  108.         local T = {}
  109.         for i = 1, #Tab do
  110.             if (Tab[i] ~= Val) then
  111.                 table.insert(T, Tab[i])
  112.             end
  113.         end
  114.         return T;
  115.     end
  116.    
  117.     -- Begin parsing
  118.     while (PARSER_POSITION < #HTMLSource) do
  119.         local CurrentChar = HTMLSource:sub(PARSER_POSITION, PARSER_POSITION)
  120.         if (DATA_STATE == "") then
  121.             if (CurrentChar == "<") then
  122.                 -- Emit text collection
  123.                 if (TEXT_COLLECTION) then
  124.                     ACTIVE_TOKEN = CreateToken("__Text")
  125.                     ACTIVE_TOKEN.Data = TEXT_COLLECTION
  126.                     ACTIVE_TOKEN.SelfClosingFlag = true
  127.                     EmitToken(ACTIVE_TOKEN)
  128.                     TEXT_COLLECTION = nil
  129.                 else
  130.                     EmitToken(ACTIVE_TOKEN)
  131.                 end
  132.                 DATA_STATE = "TAG_OPEN"
  133.                 PARSER_POSITION = PARSER_POSITION + 1
  134.             else
  135.                 -- Emit character token
  136.                 if (TEXT_COLLECTION) then
  137.                     TEXT_COLLECTION = TEXT_COLLECTION .. CurrentChar
  138.                 else
  139.                     TEXT_COLLECTION = CurrentChar
  140.                 end
  141.                 PARSER_POSITION = PARSER_POSITION + 1
  142.             end
  143.         elseif (DATA_STATE == "TAG_OPEN") then
  144.             if (CurrentChar == "!") then
  145.                 DATA_STATE = "MARKUP_DECLARATION_OPEN"
  146.                 PARSER_POSITION = PARSER_POSITION + 1
  147.             elseif (CurrentChar == "/") then
  148.                 DATA_STATE = "END_TAG_OPEN"
  149.                 PARSER_POSITION = PARSER_POSITION + 1
  150.             else
  151.                 DATA_STATE = "TAG_NAME"
  152.                 ACTIVE_TOKEN = CreateToken("")
  153.                 -- DO NOT CONSUME
  154.             end
  155.         elseif (DATA_STATE == "END_TAG_OPEN") then
  156.             if (CurrentChar == ">") then
  157.                 -- Parse error
  158.                 DATA_STATE = ""
  159.                 PARSER_POSITION = PARSER_POSITION + 1
  160.             else
  161.                 DATA_STATE = "TAG_NAME"
  162.                 ACTIVE_TOKEN = CreateToken("")
  163.                 ACTIVE_TOKEN.IsClosingTag = true
  164.                 -- DO NOT CONSUME
  165.             end
  166.         elseif (DATA_STATE == "TAG_NAME") then
  167.             if (CurrentChar == " ") then
  168.                 DATA_STATE = "BEFORE_ATTRIBUTE_NAME"
  169.                 PARSER_POSITION = PARSER_POSITION + 1
  170.             elseif (CurrentChar == "/") then
  171.                 DATA_STATE = "SELF_CLOSING_START_TAG"
  172.                 PARSER_POSITION = PARSER_POSITION + 1
  173.             elseif (CurrentChar == ">") then
  174.                 DATA_STATE = ""
  175.                 EmitToken(ACTIVE_TOKEN)
  176.                 PARSER_POSITION = PARSER_POSITION + 1
  177.             else
  178.                 ACTIVE_TOKEN.TagName = ACTIVE_TOKEN.TagName .. CurrentChar
  179.                 PARSER_POSITION = PARSER_POSITION + 1
  180.             end
  181.         elseif (DATA_STATE == "BEFORE_ATTRIBUTE_NAME") then
  182.             if (CurrentChar == " ") then
  183.                 -- Ignore
  184.                 PARSER_POSITION = PARSER_POSITION + 1
  185.             elseif (CurrentChar == "/") then
  186.                 DATA_STATE = "SELF_CLOSING_START_TAG"
  187.                 PARSER_POSITION = PARSER_POSITION + 1
  188.             elseif (CurrentChar == '"' or CurrentChar == "'" or CurrentChar == "<") then
  189.                 PARSER_POSITION = PARSER_POSITION + 1
  190.             elseif (CurrentChar == ">") then
  191.                 DATA_STATE = ""
  192.                 EmitToken(ACTIVE_TOKEN)
  193.                 PARSER_POSITION = PARSER_POSITION + 1
  194.             else
  195.                 DATA_STATE = "ATTRIBUTE_NAME"
  196.                 ACTIVE_ATTRIBUTE = AddAttribute(ACTIVE_TOKEN)
  197.                 -- DO NOT CONSUME
  198.             end
  199.         elseif (DATA_STATE == "ATTRIBUTE_NAME") then
  200.             if (CurrentChar == " ") then
  201.                 DATA_STATE = "AFTER_ATTRIBUTE_NAME"
  202.                 PARSER_POSITION = PARSER_POSITION + 1
  203.             elseif (CurrentChar == "/") then
  204.                 DATA_STATE = "SELF_CLOSING_START_TAG"
  205.                 PARSER_POSITION = PARSER_POSITION + 1
  206.             elseif (CurrentChar == '"' or CurrentChar == "'") then
  207.                 PARSER_POSITION = PARSER_POSITION + 1
  208.             elseif (CurrentChar == "=") then
  209.                 DATA_STATE = "BEFORE_ATTRIBUTE_VALUE"
  210.                 PARSER_POSITION = PARSER_POSITION + 1
  211.             elseif (CurrentChar == ">") then
  212.                 DATA_STATE = ""
  213.                 EmitToken(ACTIVE_TOKEN)
  214.                 PARSER_POSITION = PARSER_POSITION + 1
  215.             else
  216.                 PARSER_POSITION = PARSER_POSITION + 1
  217.                 ACTIVE_ATTRIBUTE.AttributeName = ACTIVE_ATTRIBUTE.AttributeName .. CurrentChar
  218.             end
  219.         elseif (DATA_STATE == "BEFORE_ATTRIBUTE_VALUE") then
  220.             if (CurrentChar == " ") then
  221.                 PARSER_POSITION = PARSER_POSITION + 1
  222.             elseif (CurrentChar == '"') then
  223.                 DATA_STATE = "ATTRIBUTE_VALUE_DOUBLE_QUOTE"
  224.                 PARSER_POSITION = PARSER_POSITION + 1
  225.             elseif (CurrentChar == '&') then
  226.                 DATA_STATE = "ATTRIBUTE_VALUE_UNQUOTED"
  227.                 -- DO NOT CONSUME THE CHARACTER
  228.             elseif (CurrentChar == "'") then
  229.                 DATA_STATE = "ATTRIBUTE_VALUE_SINGLE_QUOTE"
  230.                 PARSER_POSITION = PARSER_POSITION + 1
  231.             elseif (CurrentChar == ">") then
  232.                 DATA_STATE = ""
  233.                 EmitToken(ACTIVE_TOKEN)
  234.                 PARSER_POSITION = PARSER_POSITION + 1
  235.             elseif (CurrentChar == "<" or CurrentChar == "=") then
  236.                 PARSER_POSITION = PARSER_POSITION + 1
  237.             else
  238.                 ACTIVE_ATTRIBUTE.Value = ACTIVE_ATTRIBUTE.Value .. CurrentChar
  239.             end
  240.         elseif (DATA_STATE == "ATTRIBUTE_VALUE_DOUBLE_QUOTE") then
  241.             if (CurrentChar == '"') then
  242.                 DATA_STATE = "AFTER_ATTRIBUTE_VALUE_QUOTED"
  243.                 PARSER_POSITION = PARSER_POSITION + 1
  244.             else
  245.                 ACTIVE_ATTRIBUTE.Value = ACTIVE_ATTRIBUTE.Value .. CurrentChar
  246.                 PARSER_POSITION = PARSER_POSITION + 1
  247.             end
  248.         elseif (DATA_STATE == "ATTRIBUTE_VALUE_SINGLE_QUOTE") then
  249.             if (CurrentChar == "'") then
  250.                 DATA_STATE = "AFTER_ATTRIBUTE_VALUE_QUOTED"
  251.                 PARSER_POSITION = PARSER_POSITION + 1
  252.             else
  253.                 ACTIVE_ATTRIBUTE.Value = ACTIVE_ATTRIBUTE.Value .. CurrentChar
  254.                 PARSER_POSITION = PARSER_POSITION + 1
  255.             end
  256.         elseif (DATA_STATE == "AFTER_ATTRIBUTE_VALUE_QUOTED") then
  257.             if (CurrentChar == " ") then
  258.                 DATA_STATE = "BEFORE_ATTRIBUTE_NAME"
  259.                 PARSER_POSITION = PARSER_POSITION + 1
  260.             elseif (CurrentChar == "/") then
  261.                 DATA_STATE = "SELF_CLOSING_START_TAG"
  262.                 PARSER_POSITION = PARSER_POSITION + 1
  263.             else
  264.                 DATA_STATE = "BEFORE_ATTRIBUTE_NAME"
  265.                 -- DO NOT CONSUME CHARACTER
  266.             end
  267.         elseif (DATA_STATE == "SELF_CLOSING_START_TAG") then
  268.             if (CurrentChar == ">") then
  269.                 DATA_STATE = ""
  270.                 ACTIVE_TOKEN.SelfClosingFlag = true
  271.                 EmitToken(ACTIVE_TOKEN)
  272.                 PARSER_POSITION = PARSER_POSITION + 1
  273.             else
  274.                 DATA_STATE = "BEFORE_ATTRIBUTE_NAME"
  275.                 PARSER_POSITION = PARSER_POSITION + 1
  276.             end
  277.         elseif (DATA_STATE == "MARKUP_DECLARATION_OPEN") then
  278.             -- Are next two bytes '--' ?
  279.             if (CurrentChar == "-" and HTMLSource:sub(PARSER_POSITION + 1, PARSER_POSITION + 1) == "-") then
  280.                 DATA_STATE = "COMMENT_START"
  281.                 PARSER_POSITION = PARSER_POSITION + 2
  282.                 ACTIVE_TOKEN = CreateToken("Comment")
  283.                 -- Comments are self-closing
  284.                 ACTIVE_TOKEN.SelfClosingFlag = true
  285.             else
  286.                 DATA_STATE = "BOGUS_COMMENT"
  287.                 PARSER_POSITION = PARSER_POSITION + 1
  288.             end
  289.         elseif (DATA_STATE == "BOGUS_COMMENT") then
  290.             if (CurrentChar ~= ">") then
  291.                 PARSER_POSITION = PARSER_POSITION + 1
  292.             else
  293.                 DATA_STATE = "";
  294.                 PARSER_POSITION = PARSER_POSITION + 1
  295.             end
  296.         elseif (DATA_STATE == "COMMENT_START") then
  297.             if (CurrentChar == "-") then
  298.                 DATA_STATE = "COMMENT_START_DASH"
  299.                 PARSER_POSITION = PARSER_POSITION + 1
  300.             else
  301.                 EmitCharToOpenToken(CurrentChar)
  302.                 PARSER_POSITION = PARSER_POSITION + 1
  303.             end
  304.         elseif (DATA_STATE == "COMMENT_START_DASH") then
  305.             if (CurrentChar == "-") then
  306.                 DATA_STATE = "COMMENT_END"
  307.                 PARSER_POSITION = PARSER_POSITION + 1
  308.             elseif (CurrentChar == ">") then
  309.                 DATA_STATE = ""
  310.                 PARSER_POSITION = PARSER_POSITION + 1
  311.             else
  312.                 ACTIVE_TOKEN.Data = ACTIVE_TOKEN.Data .. CurrentChar
  313.                 PARSER_POSITION = PARSER_POSITION + 1
  314.             end
  315.         elseif (DATA_STATE == "COMMENT_END") then
  316.             if (CurrentChar == ">") then
  317.                 DATA_STATE = ""
  318.                 EmitToken(ACTIVE_TOKEN)
  319.                 PARSER_POSITION = PARSER_POSITION + 1
  320.             else
  321.                 DATA_STATE = "COMMENT_START"
  322.             end
  323.         end
  324.     end
  325.     --print("DOC_FINISH");
  326.     return DOCUMENT
  327. end
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!